High Temporal GOES Sounding Retrievals in Cloudy Regions and Applications
#Cooperative
1 Introduction
With a large possibility that the Geostationary Operational
Zhenglong Li#, Jun Li#, Paul Menzel#, James P. Nelson III#,
Timothy J. Schmit@, Elisabeth Weisz# and Steven A. Ackerman#
Institute for Meteorological Satellite Studies (CIMSS), University of Wisconsin-Madison
@Center for Satellite Applications and Research, NESDIS
Email: zhenglong.li@ssec.wisc.edu
4.2 Validation against Conventional RAOB
4.1 Validation against ARM RAOB
There are two situations where cloudy retrievals are considered as
To demonstrate that the algorithm works under different weather and
5 Application to a Severe Storm Case
The application to a tornadic storm on 24 April 2007 reveals that the
GOES cloudy sounding retrievals are more useful at the early stage of
Environmental Satellite (GOES) Sounder’s Field-Of-View (FOV) is blocked
valid: 1) When clouds are thin (thin cirrus clouds, broken low clouds and
or contaminated by clouds, it is necessary to extend the clear-sky retrievals
clouds edge), the cloudy retrievals improve the first guess in the same
to cloudy regions for better utility of sounder measurements. In this study, a
manner as clear-sky retrievals. 2) When clouds are low and thick, retrievals
sounding algorithm reveals more pronounced and extensive large
synthetic regression-based cloudy retrieval algorithm is developed and
above the cloud top are also valid. Figure 2 and 3 show the cloudy retrievals
instabilities, and it does so earlier than the clear-sky only results. Compared
applied to GOES-12 Sounder radiance measurements. This study focuses on
compared with the ARM RAOB. Figure 4 helps explain the results.
with the RUC 6-hour forecasting, the instabilities calculated from retrievals
surface conditions, validation against conventional RAOB (January 2007 to
November 2008) is also presented.
storms, when nearby clouds are considered thin or low. The GOES cloudy
are more reasonably located.
thin clouds defined as having cloud optical thickness (COT) smaller than 2.0,
and low thick clouds defined as having COT larger than 2.0 and cloud top
pressure (CTP) larger than 850 hPa.
2 Algorithm Description
The GOES single FOV (SFOV) cloudy sounding algorithm starts
thin
clouds
with a synthetic regression technique:
Y ( n ) = K ( n, m ) X ( m )
where Y is a vector of retrieval parameters (n unknowns, including cloud,
Figure 5. Validation of moisture profiles using conventional : a) relative
humidity under thin clouds, b) mixing ratio under thin clouds, c) relative
humidity under low thick clouds, d) mixing ratio under low thick clouds.
surface and atmosphere parameters), X is a vector of measurements (m
knowns, including satellite measurements and other known variables), and
K is an operator matrix to calculate Y given X. Using the least square
method, K can be obtained by
K = YX T ( XX T ) −1
3 Determination of Sounding Clouds
Figure 2. Error profiles of (a) RH and (b) 3 layer PW for thin clouds with
retrieved COT less than 2.0.
Retrievals below clouds are not reliable when clouds are optically
thick because there is little information from beneath the clouds. A
threshold COT of 2.0 has been set for two reasons. One, from Figure 1, a
smaller value decreases useful cloudy retrieval samples, although with
better retrieval precision. Secondly, a larger value introduces more retrieval
errors from low water clouds, although with more useful samples.
Low
thick
clouds
Figure 8. Time series of LI DPI imagery on 24 April 2007. From top to
bottom are 1) 20 UTC, 2) 21 UTC, 3) 22 UTC and 4) 23 UTC. From the
left to the right are a) the GFS forecast, b) the RUC 6-hour forecast, c)
GOES-12 clear-sky retrievals and d) GOES-12 cloudy retrievals.
6 Summary
With the help of NCEP GFS forecast, the regression-based cloudy
Figure 3. Same as Figure 2 except for low thick clouds with retrieved CTP
larger than 850 hPa and COT larger than 2.0. The green lines show the
clear-sky physical retrieval results with surface at the effective cloud top.
clouds
retrieval algorithm is able to retrieve the moisture profiles with reasonable
Figure 6. Validation of moisture products under the thin cloud conditions at
different TPW level using the conventional. a) the sample distribution; b)
LI; c) TPW; d) PW1; e) PW2; and f) PW3. The dark and light blue bars
represent the RMS of the forecast and the cloudy retrievals respectively.
The yellow and dark red bars represent the bias of the forecast and the
cloudy retrievals respectively.
precision in both thin clouds and low clouds conditions.. Validations against
ARM RAOB and conventional RAOB show the largest improvements in the
upper atmosphere. The middle level moisture could be improved in dry
conditions. The lower level moisture could be improved only with the help
from the surface observations in thin clouds condition. Application to a
severe tornadic case shows the cloudy retrievals provide much more
credible and earlier warnings than that from both clear retrievals and the
Figure 1. Validation of a) lifted index (LI) and b) total precipitable water
RUC 6-hour forecast to the forecasters.
(TPW) against RAOB for different cloud optical thickness. LI and TPW
7 Acknowledgement
from forecast in thin clouds are better improved than that in thick clouds;
The authors would like to thank Hal Woolf for providing the 101-level
both the RMS and the bias are reduced after the retrieval. The black solid
GOES Sounder radiative transfer model. Some data were obtained from the
line shows the sample distribution of cloud optical thickness.
Atmospheric Radiation Measurement (ARM) Program sponsored by the U.S.
4 Validation
Two different data sets are used for validation: the radiosonde (RAOB)
data from the ARM SGP site from August 2006 to May 2007 at Lamont,
OK and the conventional RAOB over the continental US from January 2007
to November 2008 . In this study, the validation focuses on moisture and
moisture related products, such as the RH profile, 3-layer PW, TPW and LI.
Figure 4. GOES-12 Sounder’s moisture weighting functions (WF) for a)
under clear skies; b) channel 10 under thin clouds; c) channel 10 under
opaque clouds. Use the US standard Atmosphere 1976. The presence of
thin cirrus clouds decreases the WF sensitivity a bit, but the three moisture
channels together are still very sensitive to PW3. Low opaque clouds
decrease the WF sensitivity a bit, but the disappearance of WF under cloud
top sharpens the WF and increases the sensitivity to PW3 and possibly
PW2.
Department of Energy, Office of Science, Office of Biological and
Environmental Research, Environmental Sciences Division. This program is
supported at CIMSS by NOAA GIMPAP program NA06NES4400002 and
Figure 7. The scatter plot of density between TPW and LI calculated from
the conventional RAOB. Notice LI is almost always greater than 0 when
TPW is smaller than 20 mm.
GOES-R program NA07EC0676. The views, opinions, and findings
contained in this report are those of the authors and should not be construed
as an official National Oceanic and Atmospheric Administration or U.S.
Government position, policy, or decision.